This is a thread for the UK-focussed OSM tools that I run at https://osm.mathmos.net/ . The tools cover a range of different things, but often involve comparing OSM data to an official dataset and flagging discrepancies for ground surveys.
I’ll try to post updates/news here when I make improvements or add new features. Feel free to post any questions or feature requests about the tools here too.
I’ve recently added a few more different brands to my “Chain Reaction” tool that looks for missing or unexpected branches of chain stores in the UK, by comparing branch lists with OSM data. The discrepancies are listed in the tool and also flagged in Survey Me!.
The data mostly comes from a great project called “All The Places” which is a collection of web spiders to extract lists of branches from company websites. These spiders sometimes break and need fixing. And there are a number of major UK chains that don’t have (working) spiders written for them yet. If anyone fancies helping out All The Places (ATP project on github), then that would be great, and it would mean I could add even more brands to my tool.
Back in October 2023, some changes to the Taginfo API meant that I coiuld no longer update the reports I was running on errors in UK Postcodes. I’m please to say that I’ve now fixed my scripts to use a different data source, and the following two reports are now working again:
OSM UK Overfull Postcode Units - lists postcodes where there are more than 100 addresses defined in OSM. (A single postcode can never contain more than 100 addresses.)
These reports both flag likely errors in OSM data, so if anyone fancies looking at them and doing some fixes where they can, that would be great. Going forward, the reports should now get updated roughly once a week with new OSM data.
First, I’ve recently added quite a few more brands to my Chain Reaction tool, which compares official branch lists to OSM data. You can look at the comparison for individual chains there, and the discrepancies are shown in my Survey Me! tool. Do have a look and see if there’s any mapping near you to be done.
Secondly, I’ve just updated the official council data in by Public Rights of Way tool for the following counties:
The city are doesn’t have parishes in the same way as most other authorities, and the numbering system is a bit messed up, because they’ve kept the original numbers when the city boundary was enlarged. But hopefully it works ok.
The PRoW Ref values need to be in the form “12” (historic City area) or “M34” (ex Meriden Rural District) or “A56” in a couple of cases (not sure what this is, but possibly Allesley Parish) possibly with a single letter suffix in a few places too. It looks like a lot of the routes have these values in a note=* tag rather prow_ref=*.
Since the prow_ref format is the same as neighbouring parts of Warwickshire, there may be some issues near the borders where my tool can’t tell which council area a way is supposed to be in. Let me know if you spot any issues (probably false positives flagged as “Way not in named parish”) and I’ll see if there’s any way to fix it.
Survey Me! shows postboxes which don’t exist any more. I assume this is because the official list comes from your 2013 FOI request, so it’s now a little out of date? Absent a new FOI request, would it possible to add a blacklist of info that has been checked and confirmed to be wrong, or do we just have to ignore it?
In my Schools Comparison Tool, I’ve just updated the list of Schools in England for the start of the new school year.
For the latter, there were many schools mapped in OSM where the ref:edubase=* number no longer exists. I’ve fixed a lot of obvious cases where the school name and location has stayed the same but governance changes have led to a new number being issued. But there are still a few hundred cases that need looking at manually. See OSM UK Schools Matching Queries if you’d like to help. Often these will be schools with mergers and/or name changes that can be verified by looking at the school websites and the data on get-information-schools.service.gov.uk .
PRoW Comparison Tool - Official data recently updated Bath and North East Somerset, Cambridgeshire, Derbyshire, Devon, East Sussex, Essex, Lincolnshire and North Yorkshire.
Chain Reaction - Data from Geolytix’s Retail Points dataset has just been updated.
Oops - thanks for spotting that. I’ve fixed once instance in my code, which will fix the occurrences on the “Access Errors” pages such as OSM Rights of Way Access Tag Errors in North Yorkshire | UK PRoW , and possibly elsewhere too. Do you know of any other occurrences that aren’t now fixed?
I’ve just updated my Public Rights of Way comparison tool for South Gloucestershire with new official data from the council. (Thanks to https://www.rowmaps.com/ for sourcing this and converting it into a standard format.)
When you update the data, do you happen to produce a diff log to list what has changed between imports? I’m thinking that could be particularly useful rather than manually trying to see what new paths have been added (by not showing up as mapped in your parish lists) or what paths have been deleted (by new prow_ref errors) by individual parish.
Perhaps you could even highlight new/edited routes in your parish lists?
Appreciate this may not be as simple as it sounds!
Do you know if the “Extinguished Rights of Way in GIS Data” table is working correctly (specifically for Lancashire)? I ask as I’m noticing some previously mapped PRoWs are showing up as “Unknown RoW Number” in your tools but are not listed in the extinguished table.
For example: Bacup FP 615 is mapped but is no longer showing up in your Bacup table, instead appearing as an unknown row number.
I couldn’t find an order extinguishing the right of way but did find an archive.org snapshot showing it used to be listed in your tables.
Hmmm, yes you’re right. Obviously when I wrote the text, it was my intention to keep records previously mapped routes that later got removed from the Definitive Map GIS data, but the current code thinks otherwise.
In order for the routes to be flagged in that section of the " Recent Changes" page, I would have to retain a RoW Record for the route, which would then mean it would appear as row in the first table of the corresponding parish page, e.g. https://osm.mathmos.net/prow/progress/lancs/rossendale/bacup/ . Is that desirable? It would also remove the OSM ways from the “Unknown RoW Number” list - although they could possibly be included there as well with a tweak to the code. (But that would mean that any manual additions for routes accidentally missing from the Definitive Map data would also appear in that list.)
I’ll need to think about this a bit more. But do you think it would be better to have the “extinguished” routes still mapped in OSM listed on the “Recent Updates” page, if that means they also have to appear in the parish table with the current rights of way? If so, should they also appear in the “Unknown RoW Number” list?